3 * Class representing a MediaWiki site.
5 * This program is free software; you can redistribute it and/or modify
6 * it under the terms of the GNU General Public License as published by
7 * the Free Software Foundation; either version 2 of the License, or
8 * (at your option) any later version.
10 * This program is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
13 * GNU General Public License for more details.
15 * You should have received a copy of the GNU General Public License along
16 * with this program; if not, write to the Free Software Foundation, Inc.,
17 * 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
18 * http://www.gnu.org/copyleft/gpl.html
22 * @license GNU GPL v2+
23 * @author John Erling Blad < jeblad@gmail.com >
24 * @author Daniel Kinzler
25 * @author Jeroen De Dauw < jeroendedauw@gmail.com >
29 * Class representing a MediaWiki site.
35 class MediaWikiSite
extends SiteObject
{
37 const PATH_FILE
= 'file_path';
38 const PATH_PAGE
= 'page_path';
43 * @param integer $globalId
45 * @return MediaWikiSite
47 public static function newFromGlobalId( $globalId ) {
48 return SitesTable
::singleton()->newRow( array(
49 'type' => Site
::TYPE_MEDIAWIKI
,
50 'global_key' => $globalId,
55 * Returns the database form of the given title.
59 * @param String $title the target page's title, in normalized form.
63 public function toDBKey( $title ) {
64 return str_replace( ' ', '_', $title );
68 * Returns the normalized form of the given page title, using the normalization rules of the given site.
69 * If the given title is a redirect, the redirect weill be resolved and the redirect target is returned.
71 * @note : This actually makes an API request to the remote site, so beware that this function is slow and depends
72 * on an external service.
74 * @note : If MW_PHPUNIT_TEST is defined, the call to the external site is skipped, and the title
75 * is normalized using the local normalization rules as implemented by the Title class.
77 * @see Site::normalizePageName
81 * @param string $pageName
86 public function normalizePageName( $pageName ) {
88 // Check if we have strings as arguments.
89 if ( !is_string( $pageName ) ) {
90 throw new MWException( '$pageName must be a string' );
93 // Go on call the external site
94 if ( defined( 'MW_PHPUNIT_TEST' ) ) {
95 // If the code is under test, don't call out to other sites, just normalize locally.
96 // Note: this may cause results to be inconsistent with the actual normalization used by the respective remote site!
98 $t = Title
::newFromText( $pageName );
99 return $t->getPrefixedText();
102 // Make sure the string is normalized into NFC (due to the bug 40017)
103 // but do nothing to the whitespaces, that should work appropriately.
104 // @see https://bugzilla.wikimedia.org/show_bug.cgi?id=40017
105 $pageName = UtfNormal
::cleanUp( $pageName );
107 // Build the args for the specific call
112 'converttitles' => true,
114 'titles' => $pageName,
115 //@todo: options for maxlag and maxage
116 // Note that maxlag will lead to a long delay before a reply is made,
117 // but that maxage can avoid the extreme delay. On the other hand
118 // maxage could be nice to use anyhow as it stops unnecessary requests.
119 // Also consider smaxage if maxage is used.
122 $url = $this->getFileUrl( 'api.php' ) . '?' . wfArrayToCgi( $args );
124 // Go on call the external site
125 //@todo: we need a good way to specify a timeout here.
126 $ret = Http
::get( $url );
129 if ( $ret === false ) {
130 wfDebugLog( "MediaWikiSite", "call to external site failed: $url" );
134 $data = FormatJson
::decode( $ret, true );
136 if ( !is_array( $data ) ) {
137 wfDebugLog( "MediaWikiSite", "call to <$url> returned bad json: " . $ret );
141 $page = static::extractPageRecord( $data, $pageName );
143 if ( isset( $page['missing'] ) ) {
144 wfDebugLog( "MediaWikiSite", "call to <$url> returned a marker for a missing page title! " . $ret );
148 if ( isset( $page['invalid'] ) ) {
149 wfDebugLog( "MediaWikiSite", "call to <$url> returned a marker for an invalid page title! " . $ret );
153 if ( !isset( $page['title'] ) ) {
154 wfDebugLog( "MediaWikiSite", "call to <$url> did not return a page title! " . $ret );
158 return $page['title'];
163 * Get normalization record for a given page title from an API response.
167 * @param array $externalData A reply from the API on a external server.
168 * @param string $pageTitle Identifies the page at the external site, needing normalization.
170 * @return array|false a 'page' structure representing the page identified by $pageTitle.
172 private static function extractPageRecord( $externalData, $pageTitle ) {
173 // If there is a special case with only one returned page
174 // we can cheat, and only return
175 // the single page in the "pages" substructure.
176 if ( isset( $externalData['query']['pages'] ) ) {
177 $pages = array_values( $externalData['query']['pages'] );
178 if ( count( $pages) === 1 ) {
182 // This is only used during internal testing, as it is assumed
183 // a more optimal (and lossfree) storage.
184 // Make initial checks and return if prerequisites are not meet.
185 if ( !is_array( $externalData ) ||
!isset( $externalData['query'] ) ) {
188 // Loop over the tree different named structures, that otherwise are similar
190 'normalized' => 'from',
191 'converted' => 'from',
192 'redirects' => 'from',
195 foreach ( $structs as $listId => $fieldId ) {
196 // Check if the substructure exist at all.
197 if ( !isset( $externalData['query'][$listId] ) ) {
200 // Filter the substructure down to what we actually are using.
201 $collectedHits = array_filter(
202 array_values( $externalData['query'][$listId] ),
203 function( $a ) use ( $fieldId, $pageTitle ) {
204 return $a[$fieldId] === $pageTitle;
207 // If still looping over normalization, conversion or redirects,
208 // then we need to keep the new page title for later rounds.
209 if ( $fieldId === 'from' && is_array( $collectedHits ) ) {
210 switch ( count( $collectedHits ) ) {
214 $pageTitle = $collectedHits[0]['to'];
220 // If on the pages structure we should prepare for returning.
221 elseif ( $fieldId === 'title' && is_array( $collectedHits ) ) {
222 switch ( count( $collectedHits ) ) {
226 return array_shift( $collectedHits );
232 // should never be here
237 * @see Site::getLinkPathType
238 * Returns Site::PATH_PAGE
244 public function getLinkPathType() {
245 return self
::PATH_PAGE
;
249 * Returns the relative page path.
255 public function getRelativePagePath() {
256 return parse_url( $this->getPath( self
::PATH_PAGE
), PHP_URL_PATH
);
260 * Returns the relative file path.
266 public function getRelativeFilePath() {
267 return parse_url( $this->getPath( self
::PATH_FILE
), PHP_URL_PATH
);
271 * Sets the relative page path.
275 * @param string $path
277 public function setPagePath( $path ) {
278 $this->setPath( self
::PATH_PAGE
, $path );
282 * Sets the relative file path.
286 * @param string $path
288 public function setFilePath( $path ) {
289 $this->setPath( self
::PATH_FILE
, $path );
293 * @see Site::getPagePath
295 * This implementation returns a URL constructed using the path returned by getLinkPath().
296 * In addition to the default behaviour implemented by SiteObject::getPageUrl(), this
297 * method converts the $pageName to DBKey-format by replacing spaces with underscores
298 * before using it in the URL.
302 * @param $pagename string: Page name (default: false)
306 public function getPageUrl( $pageName = false ) {
307 $url = $this->getLinkPath();
309 if ( $url === false ) {
313 if ( $pageName !== false ) {
314 $pageName = $this->toDBKey( trim( $pageName ) );
315 $url = str_replace( '$1', wfUrlencode( $pageName ), $url ) ;
322 * Returns the full file path (ie site url + relative file path).
323 * The path should go at the $1 marker. If the $path
324 * argument is provided, the marker will be replaced by it's value.
328 * @param string|false $path
332 public function getFileUrl( $path = false ) {
333 $filePath = $this->getPath( self
::PATH_FILE
);
335 if ( $filePath !== false ) {
336 $filePath = str_replace( '$1', $path, $filePath );